matlab-discover-clusters

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a binary MATLAB P-code file (scripts/discoverClusters.p) to perform cluster discovery. While P-code is a standard MATLAB format for protecting intellectual property, its binary nature prevents full static analysis of the logic.
  • [PROMPT_INJECTION]: The skill processes potentially untrusted metadata from network-discovered clusters, creating a surface for indirect prompt injection attacks. 1. Ingestion points: The discoverClusters function retrieves cluster names, hostnames, and properties from the local network and filesystem. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when processing discovered cluster data. 3. Capability inventory: The agent can create, modify, validate, and delete cluster profiles based on this discovered data. 4. Sanitization: The skill lacks explicit sanitization or validation of the strings returned by the discovery process before they are displayed or used in subsequent commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 12:12 PM
Security Audit — agent-trust-hub — matlab-discover-clusters